A warm lead engine that catches buying signals as they appear
Public buying signals, caught daily, judged for real intent and delivered as named people with a reason to call.
I build the software that takes the repetitive work out of your week.
Twenty five years in investment banking, private equity and as a group chief financial officer. I now design and build AI and data systems for founders, investment teams and family offices. I write the code myself, so the person you talk to on the first call is the person who delivers.
Small, concrete projects are welcome. Most of the useful work starts with one job that somebody currently does by hand every week.
Where I help
Five kinds of work
Every one of these starts as something a person currently does by hand, every week, and would rather not.
Warm leads and outreachCatching the moment somebody signals they need what you sell, while it is still worth catching.
Every day your next customer announces themselves in public. An office opening, a hiring push, a request for help with exactly the thing you do. The signals are posted in plain sight. The reason they are not leads is that nobody can read ten thousand posts a day. A machine can.
The engine scans public platforms daily, separates real intent from job advertisements and vendors and the merely curious, turns each surviving signal into a named person with a way to reach them, and delivers a ranked list with a short note on why they matter and how to open the conversation. You see the signal within days of it appearing, rather than working a list built last year.
A first projectOne segment in one market, run for a few weeks, so you can judge the quality of the leads before committing to more.
Investment and portfolio analysisPositions, prices and exposures in one place, with the risk and allocation views you actually use.
I have done this work by hand. I reviewed a portfolio running to several hundred individual asset lines across listed equity and debt, hedge funds and private equity, and built the risk measures that set what the operating companies needed in capital against what the portfolio could earn.
Doing it by hand is how you learn what matters. Building it properly is how the analysis survives the person who did it. I set up the data, the calculations and the reporting so the view refreshes without anyone rebuilding a spreadsheet.
A first projectOne consolidated view of the portfolio with the two or three measures you check most often.
Reporting, budgeting and financial controlConsolidation across entities, budget against actual, and less time assembling numbers.
As group chief financial officer I established a centralised reporting and financial control team for an industrial holding group. The platform consolidated results across every major business area and became the basis for budgeting and variance tracking.
It cut the time needed to produce the budgets by around eighty per cent, and it tracked expenditure transaction by transaction as costs emerged. That is the shape of the work: fewer hands moving numbers between files, more time spent reading what the numbers say.
A first projectOne reporting pack automated end to end, from source data to the version that goes to the board.
Documents, diligence and data roomsExtracting terms, comparing versions, and turning a data room into the questions worth asking.
I have negotiated loan documentation, hedging agreements and holding structures, and read the long documents that come with them. The reading is necessary. Doing all of it at full attention is not possible.
I build tools that pull the terms out of contracts and credit agreements, flag what has changed between versions, and summarise a data room into a list of issues. You still read the parts that matter. You stop reading the parts that do not.
A first projectA term extraction pass over one document set, checked against your own read of it, so you can see where it agrees and where it does not.
Getting a team started with AIWorking out what is worth automating, setting it up, and training the people who will use it.
Most teams do not need a strategy. They need someone to sit with them for a week, work out which parts of the job are repetitive, and set up the three things that would save real time.
I do that, write the prompts that do the job, and train the people who will run it afterwards. The measure of success is whether the team still uses it two months after I have gone.
A first projectA short review of how the week is actually spent, followed by the two changes with the clearest payback.
Proof
What this has looked like
Systems I designed and built, and what changed as a result.
A reporting platform that cut budget preparation by around eighty per cent
A single place where an industrial holding group could see its results, build its budgets and track what it was actually spending.
Reading several hundred asset lines as one portfolio
A risk and allocation review for a Forbes 100 principal, balancing what the operating businesses needed against what the financial portfolio could earn.
How it works
Three steps, and you can stop after any of them
A call
Thirty minutes. You describe the job that takes too long. I tell you whether it is worth automating, and I will say so if it is not.
A small first piece
One narrow, useful thing, delivered quickly, so you can judge the work before committing to more. Scope and price agreed in writing first.
Build and hand over
I build it, put it in front of the people who will use it, and write down how it works. You are not left depending on me to keep it running.
Tell me the job that takes too long.
Thirty minutes on a call is usually enough to know whether this is worth doing. If it is not, I will tell you that, and it will still have been a useful half hour.